Evaluation of Self-Calibrating Pressure Gauges for Seafloor Geodesy: Instrument Comparison at Axial Seamount, Barkley Canyon, and the Endeavour Segment

Seafloor quartz-resonant pressure gauges manufactured by Paroscientific have long been used to measure vertical seafloor deformation, yet the gauge-specific drift characteristics continue to hinder the precise identification of geodetic and oceanic signals. Recent self-calibrating pressure gauge designs address this limitation by housing an internal reference pressure standard that can be isolated from ambient seawater. Scheduled valve operations switch between ambient and reference pressures, thereby enabling in situ drift calibrations that are isolated from oceanographic variability and seafloor deformation. We evaluate four designs of self-calibrating pressure gauges. The University of Washington (UW) A-0-A, commercial Sonardyne Fetch AZA, and commercial RBR BPRzero use the internal pressure of the instrument housing, measured by a barometer, as a reference. In contrast, the Scripps Institution of Oceanography Cabled Self-Calibrating Pressure Recorder (CSCPR) uses a piston-cylinder system to generate a controlled reference pressure near ambient pressure reading. A-0-A, Fetch AZA, and CSCPR instruments are deployed at 1500 m depth on Axial Seamount at the Central Caldera site of the Ocean Observatories Initiative Regional Cabled Array. Additionally, a Fetch AZA is deployed at 400 m depth on the Barkley Canyon, and a BPRzero is deployed at 2200 m depth on the Endeavour segment on the Ocean Networks Canada NEPTUNE cabled observatory.

 At Axial Seamount, the instruments are within 50 m of one another and are adjacent to a conventional pressure gauge in a bottom pressure and tilt (BOTPT) instrument that has been deployed since 2014 and is well aged, with a small drift rate inferred from repeated mobile pressure recorder surveys. Assuming ocean-derived pressure fluctuations and volcanic deformation are spatially coherent across all sensors, each self-calibrating gauge is evaluated by (i) comparing its data with the BOTPT and other gauges to quantify post-calibration residuals, and (ii) assessing internal consistency for the UW A-0-A and CSCPR, which have two pressure gauges inside each unit. Our comparison shows that the self-calibrating pressure gauges generally agree within ±1.0 hPa/year (or 1.0 cm/year of water column height change) over multiple years of deployment. The one exception is the UW A-0-A system. Early in two deployments (2019-2022 and 2024-present), its two gauges are inconsistent with one another and other instruments by up to several hPa, but this bias diminishes within a year, and the records converge. We are evaluating the cause of this transient behavior by analyzing A-0-A calibration sequences. At Barkley Canyon and the Endeavour segment, we evaluate Fetch AZA and BPRzero in the same manner as at Axial Seamount, using several co-sited gauges within 3 km. Both commercial gauges agree with the independent co-sited gauge within ±1.0 hPa/year after applying drift corrections. Comparing data from co-sited sensors enables us to investigate the subtle features of each sensor's performance. All designs demonstrate the potential to reduce instrumental drift to less than 1.0 cm/year. Further evaluation across a wider range of ambient conditions and deployment configurations is warranted.